Servicenavigation

Project Details

Back

Projekt Print View

TRR 165:  Waves to Weather

Subject Area Geosciences
Computer Science, Systems and Electrical Engineering
Mathematics
Term from 2015 to 2024
Website Homepage
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 257899354
 
Our ability to predict the weather up to a week or more ahead saves our societies billions of Euros annually and protects human life and property. Exponentially increasing computing power and new observations have led to continuous improvements in forecast quality over many years, but nonetheless forecasts are sometimes strikingly poor. Increasing evidence suggests that this is not merely due to deficits in our forecasting methods, because in a chaotic atmosphere, some weather situations are intrinsically hard to predict. The great challenge today is to identify the limits of predictability in different situations and produce the best forecasts that are physically possible. The Collaborative Research Center “Waves to Weather” (CRC/Transregio 165; W2W) is conceived to meet this challenge and to deliver the underpinning science urgently needed to pave the way towards a new generation of weather forecasting systems. In Phase 2, a significant breakthrough occurred, where it was possible to show that for average conditions, the intrinsic limit of predictability will be reached when the initial condition uncertainty is reduced to about 10\% of its current level. At this point the predictability is limited by rapid error growth on small scales due to diabatic processes. In Phase 3, we will build on this knowledge by identifying windows of forecast opportunity – specific atmospheric states where extended predictions are possible. Here we will apply an innovative method using a Global Control Ensemble to investigate predicability of these relatively infrequent events. Furthermore, we will develop a Regional Grand Ensemble to obtain an unprecedented, comprehensive and quantitative view of the complex network of physical processes that contribute to rapid error growth on small scales. Finally, this theoretical knowledge will be brought to practical weather prediction by developing new, computationally efficient, representations of uncertainty with hybrid methods that combine the strengths of numerical models and machine learning. As in the previous Phases, W2W brings together the necessary broad scientific expertise from the three applying universities: the Ludwig-Maximilians-Universität (LMU) in Munich, the Johannes Gutenberg-Universität (JGU) in Mainz, and the Karlsruher Institut für Technologie (KIT). This is supported by collaboration in individual projects with the Technische Universität München (TUM), the Deutsches Zentrum für Luft- und Raumfahrt (DLR) in Oberpfaffenhofen, the Universität Hamburg (UHH) and, additionally for Phase 3, the Katholische Universität Eichstätt-Ingolstadt (KU) and the Universität Wien (UW).
DFG Programme CRC/Transregios
International Connection Austria, Netherlands

Current projects

Completed projects

 
 

Additional Information

© 2024 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung